Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
z0rlandi viSSer
Nov 5, 2013

evol262 posted:

oVirt has both of these features, plus HA and live migration.

Thanks, I will look into this more. We spend a lot of money of red hat i don't know why we don't use RHEV.

Adbot
ADBOT LOVES YOU

EuphrosyneD
Jan 25, 2004
Probably a dumb question to ask, but I wanted to know about a suitable lab environment for earning the MCSE 2012: Messaging certification. I keep finding not knowing Exchange to be a pain point in my employment.

At the very least (I think), this would require:

1 VM for Windows Server 2012 to serve as my domain controller/DNS/DHCP server - Server Core install?
1 VM for Windows Server 2012 to serve as my Exchange box
Space for 2-3 other VMs to be used as needed

The goals of this would be to be able to simulate a small office that can send and receive email from the Internet, similar to our clients.

Hardware would be a recent Intel NUC (or two for clustering?) with as much RAM as I can get in there, or if that's recommended against, maxing out the RAM and doing a SSD upgrade on my current home PC, and a cheapish NAS or SAN appliance for network storage. I don't really want to spend $bignums on this, and I would prefer a quiet and low-power host as space in my apartment is limited.

The software I can borrow from work (my boss gave the OK to download from our Microsoft partner webpage).

I was thinking about ESXi as my hypervisor.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

EuphrosyneD posted:

Probably a dumb question to ask, but I wanted to know about a suitable lab environment for earning the MCSE 2012: Messaging certification. I keep finding not knowing Exchange to be a pain point in my employment.

At the very least (I think), this would require:

1 VM for Windows Server 2012 to serve as my domain controller/DNS/DHCP server - Server Core install?
1 VM for Windows Server 2012 to serve as my Exchange box
Space for 2-3 other VMs to be used as needed

The goals of this would be to be able to simulate a small office that can send and receive email from the Internet, similar to our clients.

Hardware would be a recent Intel NUC (or two for clustering?) with as much RAM as I can get in there, or if that's recommended against, maxing out the RAM and doing a SSD upgrade on my current home PC, and a cheapish NAS or SAN appliance for network storage. I don't really want to spend $bignums on this, and I would prefer a quiet and low-power host as space in my apartment is limited.

The software I can borrow from work (my boss gave the OK to download from our Microsoft partner webpage).

I was thinking about ESXi as my hypervisor.



https://www.youtube.com/watch?v=Ntf0SBmWL1U

This goes over pretty indepth about how to build setup and run a lab.


Wih that a simple dual core +HT, 16 GB ram, and a 128GB SSD/1TB HDD would probably be enough.

evol262
Nov 30, 2010
#!/usr/bin/perl

dont change my name posted:

Thanks, I will look into this more. We spend a lot of money of red hat i don't know why we don't use RHEV.

RHEV is a separate entitlement, so not actually free. oVirt's upstream, though. The primary differences between RHEV and oVirt are integration with RHN, RHEV being red instead of blue, and being able to get a TAM or GSS rep to harass me for problems instead of doing it yourself on oftc/#ovirt

There are some extremely large core partners (~6k VMs at one), but we're still way smaller than vmware, and the ecosystem shows it. But it's by far the best feature set you'll get for free.

Read the wiki page about hosted engine, though. In the past, the management engine couldn't run on your environment in case of dependency issues during a hard crash, etc. Better now, but it's very new, and I'm not sure what manual steps are needed (if any; it should have made it into 3.4). Better safe

Mniot
May 22, 2003
Not the one you know
Anyone have experience with VirtualBox and VMWare Workstation inside of vSphere?

I'm using Vagrant to hold a development testing environment. On my desktop machine, it works just fine with VirtualBox. But I'd like to have the office Jenkins machine run CI tests with it, and Jenkins is living on an ESXi VM.

I've tried Vagrant+VirtualBox inside the VM, and performance is too slow. Some system commands time out in weird ways and the whole thing takes at least 40 minutes to run (vs 2 minutes on my laptop).

Vagrant can use VMWare instead of VirtualBox. It requires paying for a license, so I'm wondering if there's some evidence that it would be worth it. Has anyone run different VMs under ESXi and noticed a significant difference? Is my problem just that the guest machine is underpowered?

evol262
Nov 30, 2010
#!/usr/bin/perl

Mniot posted:

Anyone have experience with VirtualBox and VMWare Workstation inside of vSphere?

I'm using Vagrant to hold a development testing environment. On my desktop machine, it works just fine with VirtualBox. But I'd like to have the office Jenkins machine run CI tests with it, and Jenkins is living on an ESXi VM.

I've tried Vagrant+VirtualBox inside the VM, and performance is too slow. Some system commands time out in weird ways and the whole thing takes at least 40 minutes to run (vs 2 minutes on my laptop).

Vagrant can use VMWare instead of VirtualBox. It requires paying for a license, so I'm wondering if there's some evidence that it would be worth it. Has anyone run different VMs under ESXi and noticed a significant difference? Is my problem just that the guest machine is underpowered?

Vagrant is significantly more mature with Virtualbox. VMware support is very new, as is KVM support. Do you have nested virt enabled? Because if you don't, virtualbox will default to (slow, awful) binary translation.

Mniot
May 22, 2003
Not the one you know

evol262 posted:

Vagrant is significantly more mature with Virtualbox. VMware support is very new, as is KVM support. Do you have nested virt enabled? Because if you don't, virtualbox will default to (slow, awful) binary translation.

I'm new to VMs and the IT staff I'm working with are all Windows-only so there's a lot of fumbling around.

Is this nested virtualization a setting on VirtualBox? When I initially tried to do this, VB just refused to boot anything. It turned out that I hadn't turned on "hardware virtualization" in vSphere for the host VM. That's the only switch I've flipped.

luminalflux
May 27, 2005



Sounds like Docker is a closer fit than Virtualbox or Workstation for your case.

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.
e: reading helps discern context tremendously!

Vulture Culture fucked around with this message at 18:05 on Mar 3, 2014

Docjowles
Apr 9, 2009

Am I understanding right that you have ESXi, which is running a Linux VM, which in turns runs Virtualbox, which in turn spins up its own VM's via Vagrant? That's a whole lot of layers of indirection. Have you tried the vagrant-esxi plugin? That would let you remove the Virtualbox layer and have your Jenkins builds spin up new VM's directly inside ESXi. I should add that I have no experience with that particular plugin since I don't work in a VMware environment. It may be buggy as hell for all I know :haw: But I'm actually doing something similar with the vagrant openstack plugin and it works well. "vagrant up" creates a new KVM VM inside our OpenStack cloud and it owns.

edit: Or possibly vagrant-vsphere. Seems to be better documented and support more config options for where and how to place the VM.

Docjowles fucked around with this message at 18:08 on Mar 3, 2014

evol262
Nov 30, 2010
#!/usr/bin/perl

luminalflux posted:

Sounds like Docker is a closer fit than Virtualbox or Workstation for your case.

Docker is extraordinarily limited in many ways, and may not be suitable, though flat LXC may be, depending on what's being done with Vagrant.

Docjowles posted:

Am I understanding right that you have ESXi, which is running a Linux VM, which in turns runs Virtualbox, which in turn spins up its own VM's via Vagrant? That's a whole lot of layers of indirection. Have you tried the vagrant-esxi plugin? That would let you remove the Virtualbox layer and have your Jenkins builds spin up new VM's directly inside ESXi. I should add that I have no experience with that particular plugin since I don't work in a VMware environment. It may be buggy as hell for all I know :haw: But I'm actually doing something similar with the vagrant openstack plugin and it works well. "vagrant up" creates a new KVM VM inside our OpenStack cloud and it owns.

edit: Or possibly vagrant-vsphere. Seems to be better documented and support more config options for where and how to place the VM.

vagrant-vsphere and vagrant-kvm are awesome if you don't mind rebundling the box, but this is a hassle for some significant proportion of vagrant users, and most of the more advanced (teaming, shared folders, etc) features only sort-of worked on vsphere or kvm.

This is the first I've even heard of vagrant-openstack. It looks neat, but I'm really curious what advantages it offers over just using the openstack tools directly. It looks like it doesn't do anything that a userdata script to cloud-init doesn't already do (dummy boxes that use openstack volmes, etc). Am I missing something? It looks like it'd be great if I didn't have to specify the floating IP manually.

Docjowles
Apr 9, 2009

Mostly just integration with existing work flows. Devs are used to typing "vagrant up", but maybe they need a VM that's beefier than what their own laptop can handle. Or one that a bunch of people are going to access collaboratively that needs to be longer lived. They don't really need to do anything different to get one. Just give them a new Vagrantfile, "vagrant up" and they have a VM.

We don't use floating IP's so I can't speak to that. Our openstack setup is pretty simple.

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.

evol262 posted:

Docker is extraordinarily limited in many ways, and may not be suitable, though flat LXC may be, depending on what's being done with Vagrant.
vagrant-lxc is pretty neat.

Mniot
May 22, 2003
Not the one you know

luminalflux posted:

Sounds like Docker is a closer fit than Virtualbox or Workstation for your case.

I'd love to have our stuff in Docker, but we've got a tangled mess of applications right now. I think we'd need to build a container for each one, do some dependencies, and get them to talk. It would be superior in basically every way, but I can't burn that much development time making invisible improvements. Stuffing everything into Vagrant took a couple of hours and only involved writing some shell scripts and some Ruby.

Docjowles posted:

Am I understanding right that you have ESXi, which is running a Linux VM, which in turns runs Virtualbox, which in turn spins up its own VM's via Vagrant?

No, I was unclear.

ESXi, running a Linux VM. Then the Linux VM uses Vagrant to spin up its own VM. When Vagrant uses VirtualBox, performance sucks, and so I'm wondering if using the VMWare provider would help (because it's VMWare inside VMWare?).

quote:

edit: Or possibly vagrant-vsphere. Seems to be better documented and support more config options for where and how to place the VM.

Maybe you're right. I'd seen vagrant-esxi but the documentation seemed sparse and I have to convince IT to give me some kind of super-user powers on their vSphere to do Linux stuff that they don't understand. (They're nice guys but it's an awkward position for them.) Still, it's worth a hassle to solve the problem correctly.

evol262
Nov 30, 2010
#!/usr/bin/perl

Mniot posted:

I'd love to have our stuff in Docker, but we've got a tangled mess of applications right now. I think we'd need to build a container for each one, do some dependencies, and get them to talk. It would be superior in basically every way, but I can't burn that much development time making invisible improvements. Stuffing everything into Vagrant took a couple of hours and only involved writing some shell scripts and some Ruby.


No, I was unclear.

ESXi, running a Linux VM. Then the Linux VM uses Vagrant to spin up its own VM. When Vagrant uses VirtualBox, performance sucks, and so I'm wondering if using the VMWare provider would help (because it's VMWare inside VMWare?).


Maybe you're right. I'd seen vagrant-esxi but the documentation seemed sparse and I have to convince IT to give me some kind of super-user powers on their vSphere to do Linux stuff that they don't understand. (They're nice guys but it's an awkward position for them.) Still, it's worth a hassle to solve the problem correctly.

Enable nested virt.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

Mniot posted:

ESXi, running a Linux VM. Then the Linux VM uses Vagrant to spin up its own VM. When Vagrant uses VirtualBox, performance sucks, and so I'm wondering if using the VMWare provider would help (because it's VMWare inside VMWare?).

Enable nested virtualization on the VM,

if 5.1 ssh into your esxi box
goto
/vmfs/volumes/<datastorename>/$vm$/%vm_name%.vmx

add a line

VHV.enable = "TRUE"

write and quit

and make sure to use hardware acceloration, it's in the advanced tab under virtual machine settings. Don't have it up right now so I'm taking from memory.

fletcher
Jun 27, 2003

ken park is my favorite movie

Cybernetic Crumb
Having some trouble with shared folders in Virtual Box. I've used these a bunch before and never had a problem.

Host OS is Windows 7, guest OS is Linux Mint 16 MATE (kernel is 3.11.0-12-generic). Using the VirtualBox, 4.3.8.

Created the shared folder as normal, auto-mount & make permanent checked. Added my linux user to the vboxsf group. I can see the shared folder in /media now, but if I try to ls it, both CPU cores get pegged @ 100% and it just kinda hangs.

I feel like I ran into this a couple months ago but I can't remember what I did to fix it. :(

Found some similar looking bug reports: https://bugzilla.novell.com/show_bug.cgi?id=841673 & https://www.virtualbox.org/ticket/11947

edit: Force installing the guest additions seems to have fixed it, I had to manually overwrite the version that mint came with:
code:
fletcher@linuxmint16 /media/fletcher/VBOXADDITIONS_4.3.8_92456 $ sudo ./VBoxLinuxAdditions.run
Verifying archive integrity... All good.
Uncompressing VirtualBox 4.3.8 Guest Additions for Linux............
VirtualBox Guest Additions installer
You appear to have a version of the VBoxGuestAdditions software
on your system which was installed from a different source or using a
different type of installer.  If you installed it from a package from your
Linux distribution or if it is a default part of the system then we strongly
recommend that you cancel this installation and remove it properly before
installing this version.  If this is simply an older or a damaged
installation you may safely proceed.

Do you wish to continue anyway? [yes or no]

fletcher fucked around with this message at 20:27 on Mar 4, 2014

Erwin
Feb 17, 2006

I have a question about VirtualBox vs VMware Fusion on OS X. I have a new rMBP and Fusion makes for an awesome experience - it supports the fullscreen space dealy, and the Windows 8.1 guest is aware of the retina resolution and looks fantastic. Virtualbox VMs however are laggy as hell and not resolution aware. Googling isn't much help - is there something I'm missing with VirtualBox? Or is that just where it's currently at?

In that case, I don't mind paying for Fusion (my trial is up), but since my company pays VMware buttloads of money already, is there like a deal for vSphere licensees?

Docjowles
Apr 9, 2009

Erwin posted:

I have a question about VirtualBox vs VMware Fusion on OS X. I have a new rMBP and Fusion makes for an awesome experience - it supports the fullscreen space dealy, and the Windows 8.1 guest is aware of the retina resolution and looks fantastic. Virtualbox VMs however are laggy as hell and not resolution aware. Googling isn't much help - is there something I'm missing with VirtualBox? Or is that just where it's currently at?

In that case, I don't mind paying for Fusion (my trial is up), but since my company pays VMware buttloads of money already, is there like a deal for vSphere licensees?

Probably obvious but have you installed the guest extensions for VBox? Comparable to VMware Tools in the vSphere world. Also go into the VM's properties in VirtualBox and make sure hardware virtualization is enabled under System > Acceleration.

I also experienced abysmal networking performance til I started using the Paravirtualized network adapter.

CtrlMagicDel
Nov 11, 2011
I'm really struggling with the performance charts in vCenter vs. what the guest OS is reporting for memory usage. I have an 8GB Windows VM that when reviewing the vCenter usage metric for Memory appears to fluctuate between 15-25%. However looking at the guest OS it appears to be 80% utilized, and I can see 3 1GB processes and a 2GB process running, along with all of the usual Windows processes so it definitely appears to be using well more than 15-25% of memory. When I took the Optimize and Scale class the instructor really tried to drive home "Don't trust the guest OS it is lying to you only trust vCenter!" but I'm currently staring at a VM where these metrics don't seem to agree at all. Is there a better metric than usage I should be using in vCenter? Nothing is swapped out according to vCenter.

I'd also be curious to know how other people judge when a VM needs more space.

Erwin
Feb 17, 2006

Are you just looking at the overview section of the performance tab of the VM? That shows active memory, which is estimated by the hypervisor and is almost always lower than what's actually consumed. Use an advanced chart and look at the consumed metric. It'll be closer to what's granted to the VM, depending on transparent page sharing and the like. The guest OS's metrics don't actively reflect the physical memory used on the hypervisor (8GB does not necessarily mean it's using 8 of the 64GB or whatever on the host) but its shown percentage in the OS is the percentage it is using of what's been made available to it.

If the VM is paging to its disk, it's time to give it more. You may not want to give it more at 80%, though, if it's not paging. As it uses memory, the hypervisor will allocate physical memory to it, but it doesn't know when it has freed memory, so it can't actually free up the physical memory. Let's say it's using 6 of the 8GB it has. You give it 12GB, and through starting and stopping applications and whatever it does, it ends up writing to - and then freeing, most of that 12GB. It could be back down to 6GB used, but the hypervisor has it assigned to say 10GB of host memory. If it's out of physical memory, the hypervisor will inflate the balloon driver, and wherever it sees that appear in physical memory, it knows that the OS is no longer using that and can overwrite it with pages from another VM.

Check out this document: http://www.vmware.com/files/pdf/mem_mgmt_perf_vsphere5.pdf

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug
Any good books for Citrix VDI advanced learning? I mean I can set it up and administer it; but I want to dive more into optimizing and scaling on the level I know vmware view.

KennyG
Oct 22, 2002
Here to blow my own horn.
Can anyone help me with:

VMHost posted:

Configuration Issues
No vmkcore disk partition is available and no network coredump server has been configured. Host core dumps cannot be saved.

I have 6 hosts in a cluster and 2 of the 6 display the message. The other 4 don't. I didn't do anything fancy to configure them. Can anyone explain what is going on here or how to fix it. I don't think the google diagnosis is what I should do as it's a rather involved SSH/CLI solution that sends me down a rabbit hole I didn't do for the other 4.

CtrlMagicDel
Nov 11, 2011

Erwin posted:

Are you just looking at the overview section of the performance tab of the VM? That shows active memory, which is estimated by the hypervisor and is almost always lower than what's actually consumed. Use an advanced chart and look at the consumed metric. It'll be closer to what's granted to the VM, depending on transparent page sharing and the like. The guest OS's metrics don't actively reflect the physical memory used on the hypervisor (8GB does not necessarily mean it's using 8 of the 64GB or whatever on the host) but its shown percentage in the OS is the percentage it is using of what's been made available to it.

If the VM is paging to its disk, it's time to give it more. You may not want to give it more at 80%, though, if it's not paging. As it uses memory, the hypervisor will allocate physical memory to it, but it doesn't know when it has freed memory, so it can't actually free up the physical memory. Let's say it's using 6 of the 8GB it has. You give it 12GB, and through starting and stopping applications and whatever it does, it ends up writing to - and then freeing, most of that 12GB. It could be back down to 6GB used, but the hypervisor has it assigned to say 10GB of host memory. If it's out of physical memory, the hypervisor will inflate the balloon driver, and wherever it sees that appear in physical memory, it knows that the OS is no longer using that and can overwrite it with pages from another VM.

Check out this document: http://www.vmware.com/files/pdf/mem_mgmt_perf_vsphere5.pdf

Man that is some dense material to read on a Friday.

So one thing to mention is that hosts in our environment is not overcommitted on memory and not really anywhere close to 100% memory utilization. So looking as the Consumed memory metric under the advanced charts it is basically always a flat line roughly equal to the amount of memory the VM is allocated with since no memory reclamation except maybe some small amount of transparent page sharing is taking place? The active chart fluctuates pretty wildly, though looking from the guest OS in task manager the memory usage is generally a lot more stable. In this sort of environment does it just make more sense to rely on the guest OS measurements since it is probably pretty accurate? The consumed and active memory measurements really don't seem to provide much value.

CtrlMagicDel
Nov 11, 2011

KennyG posted:

Can anyone help me with:


I have 6 hosts in a cluster and 2 of the 6 display the message. The other 4 don't. I didn't do anything fancy to configure them. Can anyone explain what is going on here or how to fix it. I don't think the google diagnosis is what I should do as it's a rather involved SSH/CLI solution that sends me down a rabbit hole I didn't do for the other 4.

Do you configure any sort of network core dumping?
Do those 2 hosts not have internal hard drives and the other ones do?

RusteJuxx
Jul 14, 2001

College Slice
Right now I'm running HyperV on a Server 2008 host that I have 8 VMs running on it. The warranty on the server is expiring at the end of June, so that means time for a new server. HyperV has been rock solid for the past five years, but I want to roll out VMware's products so that I have a better understanding of them for when I move on from this current position. I can't find a clear cut response via Google to the few questions I have about that I was hoping to get some input on.

1. With HyperV I have Microsoft Data Protection Manager 2007 performing my live nightly backups of the machines and Tivoli is running on the machines for file level backup. What product from VMware handles live backups? I see tons of information about live migrations, fall back, and disaster recovery, but on a basic level - which package do I definitely need in order to have this functionality? Is there anything that provides file level backup and restoration, too? I'd love to be able to forgo Tivoli on some of them as its performance isn't great.

2. What's the goon recommended process for performing live migrations to VMware? Is there a process or application built in the VMware's suite to perform this function or is easiest to just the Converter application to change between them? Our datacenter is terrible and the current server is on 100 Mb connection, so it's going to be painful to do the transfers and I definitely don't want to worry about it failing and needing to do it again.

3. Not exactly virtualization related, but while I'm asking questions... is there an enterprise hardware thread I'm just overlooking? I see some specific things, such as the enterprise networking and data, but I was hoping to get someone to weigh in on the specs of my new server build before committing.

Thank you!

complex
Sep 16, 2003

KennyG posted:

I have 6 hosts in a cluster and 2 of the 6 display the message. The other 4 don't. I didn't do anything fancy to configure them. Can anyone explain what is going on here or how to fix it. I don't think the google diagnosis is what I should do as it's a rather involved SSH/CLI solution that sends me down a rabbit hole I didn't do for the other 4.

Check that your local drives have a diagnostic partition. They likely don't. If they are disk builds, make sure you installed to the right disk, didn't it get confused with on board SD card or something else. If you are using Auto Deploy ensure that the local disks either have a diagnostic partition already, or are writable so one can be created. Use partedUtil to explore/wipe if necessary.

EMILY BLUNTS
Jan 1, 2005

Your VMWare servers are running smoothly, but you're awake at night. You just know you're getting a call, or an alert, about the server, and soon. What's about to go wrong?

Bhodi
Dec 9, 2007

Oh, it's just a cat.
Pillbug
One of your datastores has filled up and hung every virtual server on it. Capacity planning? Budget? Monitoring? Nah. It's kind of like gas, there are just some people who will run it past empty and then act surprised when the car stops in the middle of the highway. At least I don't own it!

It's happened twice in the last month :negative:

Kachunkachunk
Jun 6, 2011
Isn't thin on thin on thin so simple to manage? :smithicide:

El_Matarife
Sep 28, 2002
I'm running a study group tomorrow night for the VCAP5-DCA. We're going to start by covering the three objectives very few people use in real life: Autodeploy, Kickstart files, and Private VLANs. I'm using Jason Langer's VCAP5-DCA study guide http://www.virtuallanger.com/vcap-dca-5/ and Chris Wahl's DCA study guide http://wahlnetwork.com/2012/07/02/the-vcap5-dca-study-sheet/&#65279; plus a hands on lab that covers some autodeploy. Anyone have any other good suggestions? I was going to recommend readings in Scott Lowe's Mastering vSphere 5 book too.

(And if you want to attend, PM me.)

Winkle-Daddy
Mar 10, 2007
Anyone have experience with kitchen-docker? I'm making a bare bones image and every time I build it and try to login I get connection closed by remote host. Both ssh'ing to root@localhost -p [docker port] as well as using kitchen login [image name]. I'm wondering if it might be SELinux on Fedora 20 and building a centos image something stupid. I'm building an Ubuntu test right now to see if that fixes the issue, but if anyone's accomplished this, I'd love to know about it.

e: I know the password is good, wrong password results in a different error.

Winkle-Daddy fucked around with this message at 03:41 on Mar 11, 2014

Docjowles
Apr 9, 2009

I'll say up front I have minimal docker experience, but usually when SSH just insta fails with an error like that it's a networking problem in my experience. It's not even getting far enough into the handshake to ask for a password. Double check that the host and the container have the network settings you expect.

Failing that, I agree that SElinux is always a good bet when poo poo is failing for no god damned reason.

GrandMaster
Aug 15, 2004
laidback
Anyone well versed in VMs inside VMs?
I'm trying to set up a temporary environment here..

ESXi 4.0 nested inside ESXi 5.1
This is for a Lab Manager environment, the hardware has to go so I want to temporarily set up the ESXi 4.0 guest VM, and present the existing datastore to it as an RDM. Lab Manager requires 4.0/4.1 and doesn't work with 5.x, hence the VM.

Now the problem is, that when I connect the datastore LUNs to the VM cluster, the host picks it up as VMFS so it's not available for me to connect to the guest esx as an RDM.

I've tried unmounting and creating the RDM file via the CLI but it's saying it can't lock the file so now I'm kinda stumped.
Another solution would be to just create new LUNs, present them as RDM, create the VMFS from within the guest esx and migrate the data onto them, but the lab manager file migration tool is super slow, and an offline operation.
They are FC, so it has to be done via RDM - iSCSI direct to the guest would be handy, but our array has no IP interfaces/

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

RusteJuxx posted:

Right now I'm running HyperV on a Server 2008 host that I have 8 VMs running on it. The warranty on the server is expiring at the end of June, so that means time for a new server. HyperV has been rock solid for the past five years, but I want to roll out VMware's products so that I have a better understanding of them for when I move on from this current position. I can't find a clear cut response via Google to the few questions I have about that I was hoping to get some input on.
Don't play test bed on your production environment; if hyper-v is working fine use it, converting is going to cause downtime. If you are single host environment and already running happy with Hyper-V what other than personal knowledge gains are you getting? I mean I do feel ESXi could offer a bit better performance in a few area's, but how strained are you currently?

quote:

1. With HyperV I have Microsoft Data Protection Manager 2007 performing my live nightly backups of the machines and Tivoli is running on the machines for file level backup. What product from VMware handles live backups? I see tons of information about live migrations, fall back, and disaster recovery, but on a basic level - which package do I definitely need in order to have this functionality? Is there anything that provides file level backup and restoration, too? I'd love to be able to forgo Tivoli on some of them as its performance isn't great.
Veeam has a free one, VDP is "free" with essentials plus, works great. Most virtual backup software leverage snapshots on the VM, then commits it to a replica VM or the like.

quote:

2. What's the goon recommended process for performing live migrations to VMware? Is there a process or application built in the VMware's suite to perform this function or is easiest to just the Converter application to change between them? Our datacenter is terrible and the current server is on 100 Mb connection, so it's going to be painful to do the transfers and I definitely don't want to worry about it failing and needing to do it again.
P2V's work great but will cause a slight outage similar to an HA event. Expect about 10-15 minutes as machines boot up on the target.
Usually you get a 1GB/min on P2V's on a gig network.

quote:

3. Not exactly virtualization related, but while I'm asking questions... is there an enterprise hardware thread I'm just overlooking? I see some specific things, such as the enterprise networking and data, but I was hoping to get someone to weigh in on the specs of my new server build before committing.

Dunno but ask away, most all brand name enterprise gear works fine.

GrandMaster posted:

Anyone well versed in VMs inside VMs?
I'm trying to set up a temporary environment here..

ESXi 4.0 nested inside ESXi 5.1
This is for a Lab Manager environment, the hardware has to go so I want to temporarily set up the ESXi 4.0 guest VM, and present the existing datastore to it as an RDM. Lab Manager requires 4.0/4.1 and doesn't work with 5.x, hence the VM.

Now the problem is, that when I connect the datastore LUNs to the VM cluster, the host picks it up as VMFS so it's not available for me to connect to the guest esx as an RDM.

I've tried unmounting and creating the RDM file via the CLI but it's saying it can't lock the file so now I'm kinda stumped.
Another solution would be to just create new LUNs, present them as RDM, create the VMFS from within the guest esx and migrate the data onto them, but the lab manager file migration tool is super slow, and an offline operation.
They are FC, so it has to be done via RDM - iSCSI direct to the guest would be handy, but our array has no IP interfaces/

I do it a bunch for my college, but I use IP for it all. with doing FC your safest/best bet is probably local VMDK on the host or doing a device passthough on a single port FC adapter. Problem is you lose HA on that VM.

Dilbert As FUCK fucked around with this message at 06:36 on Mar 11, 2014

GrandMaster
Aug 15, 2004
laidback
Local VMDK isn't an option - The data needs to be shared between a physical node ESXi4 during the migration.
Was looking at NPIV to zone the LUN directly to the VM, but it seems that brocade gouge you for licensing on every single feature you want to enable :(

Passthrough of a single port looks to be the best option, i'll lose my MPIO but since it's temporary it shouldn't matter for the short term.

EMILY BLUNTS
Jan 1, 2005

VMware is it so hard to:
Make a vSphere vCenter vConverter that vSupports vAll the vVersions of ESXi/Workstation and can do physical disk->vmdk?
p2v is the foot in the door, of all the tools this one should make any crazy combo of source/destination relatively painless.

Making a vm so I can attach the disk and then use convertor to send it into the server is... a bit silly. The alternative is to hunt down 7 year old versions of imaging software.

Maybe I missed out on some 3rd party thing or a super-hidden VMware tool.

Mausi
Apr 11, 2006

Sooo, a client uses a non-routed subnet to connect their NFS vmkernel ports to their NetApp, and mounts datastores using IP:/ExportPath. All datastores are connected to all hosts to make migrating VMs around simple, so far so good.
The problem is that this subnet is now nearly full, so how to get them more IPs for their new hosts when they can't simply expand the subnet because other services have the surrounding ranges? This is their Production environment, so simply turning everything off and switching to a new subnet isn't really an option.

I'm aware that vCenter throws a snit if the datastore mount path isn't identical, so the best plan I've thought of so far is to commision a new larger subnet, hook it up in parallel to the existing configuration, provision a bunch of new storage and write up some svMotion scripts.
The problem with this approach is that the environment in question is approximately 1300 vmdk's totally about 40Tb (thin that is, 110Tb allocated), and they're not keen on such things running during the day.

Anybody else got a better solution?

evil_bunnY
Apr 2, 2003

Lower subnet mask, profit?

But seriously, good luck :|

Adbot
ADBOT LOVES YOU

Bob Morales
Aug 18, 2006


Just wear the fucking mask, Bob

I don't care how many people I probably infected with COVID-19 while refusing to wear a mask, my comfort is far more important than the health and safety of everyone around me!

What are you guys doing in the case of "virutalize a pile of old PC's running random poo poo"

I'm thinking we can just buy some Win7 Pro licenses and run the apps in compatibility mode, since they are all XP right now.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply